Colloque : Nouvelles approches pour le suivi des plantes, des pollinisateurs et de leurs interactions dans un monde en changement
Conférence du 23 mai 2024 : The Spipoll Project: Monitoring Plant-Visitor Interactions in France with Citizen Science
Intervenant :
Nicolas Deguines, Maître de conférences, Écologie et Biologie des Interactions, université de Poitiers, Poitiers, France
In the context of global changes affecting biodiversity worldwide, to what extent flower visitors —and not just bees— are threatened by different environmental factors needs to be evaluated. This is critical if we are to mitigate the effects of global changes on these organisms along with their functions in ecosystems and the services they provide to human societies.
In 2010, the Photographic Survey of Flower Visitors (Spipoll) was launched to monitor plant-visitor interactions across France with the help of citizen scientists following a standardized protocol. During 20 minutes exactly, a volunteer takes pictures of every organisms visiting any flower of a chosen plant species. Date, time, location and conditions of observations, along with identifications of the plant and its visitors are provided by volunteers to a taxonomic resolution as fine as possible based on morphological characteristics visible on photographs (and using an interactive identification key specifically developed for the Spipoll). In this talk, after introducing the Spipoll protocol and unique dataset collected (more than 650,000 plant-visitor interactions across 75,000+ sampling locations), we will present the main results obtained since 2010 and develop scientific perspectives.
Retrouvez les enregistrements audios et vidéos du cycle :
https://www.college-de-france.fr/fr/agenda/colloque/nouvelles-approches-pour-le-suivi-des-plantes-des-pollinisateurs-et-de-leurs-interactions-dans-un
Chaire Biodiversité et écosystèmes
Professeure : Emmanuelle Porcher
Retrouvez tous ses enseignements :
https://www.college-de-france.fr/fr/chaire/emmanuelle-porcher-biodiversite-et-ecosystemes-chaire-annuelle
Le Collège de France est une institution de recherche fondamentale dans tous les domaines de la connaissance et un lieu de diffusion du « savoir en train de se faire » ouvert à tous.
Les cours, séminaires, colloques sont enregistrés puis mis à disposition du public sur le site internet du Collège de France.
Découvrez toutes les ressources du Collège de France :
https://www.college-de-france.fr
Soutenir le Collège de France :
https://www.fondation-cdf.fr/faire-un-don
Suivez-nous sur :
Facebook : https://www.facebook.com/College.de.France
Instagram : https://www.instagram.com/collegedefrance
X (ex-Twitter) : https://x.com/cdf1530
LinkedIn : https://fr.linkedin.com/company/collègedefrance
[Music] [Music] okay next speaker is Nicola duin from the University of pu good afternoon thank you Gabrielle and emuel for inviting me to discuss about the spe project so I’m going to present you the project and his main its main results uh before to start I’d like to an acknowledge that it’s not just me talking it’s all the collaborators on the project so first the participant of the people uh the op which is an anical NGO so in particular my Flores and Pascal dupon and Roman J K Fon Greg Lis and many others um so I’m going to start with the objectives of the people and then present you how it works exactly and a little bit of what changed since the beginning of the program I will show you the main results and some scientific perspectives uh when we talk about pollinators often the first thing that comes to mind uh for the public but as Dan showed also for scientist or in the literature we are thinking about honey bees bumblebees and maybe sometimes wild bees there are reasons for that there are many species of bees worldwide and both adult saral stages feed on nectar and pollen so that makes them um very good pollinators and uh so most pollination ecologists would agree that bees are a predominant pollinator for plants and ecosystems but bees are not alone out there and this study for example was looking at pollinator frequency by groups and for different plant species so in this pie chart we have the visitation frequency for different groups and B here it’s only about a third of the visits to this plant and most of the visits comes from non surit flies and then comes other groups as well if we look at other plant species that it it will be a different set of visitors in different proportions and the main message is that each plant is visited by by a high diversity of flower visitors and it’s actually very difficult to know precisely for sure which is pollinating and which is just visiting without pollinating and some Works have shown or conclude that the frequency of visits and flower can be used as a surrogate for efficiency and even if there may be some um um maybe there is some time it’s not working but at least it seems important to consider to consider all flower visitors so it’s not just bees it’s also other imoan such as ants wasps and slies it’s a lot of flies butterflies and mffs beetles and even spiders there are a lot of scrab crab spiders for example on the flowers and with the spe project we look we look at we include all of these groups uh all these groups have have very different life traits in terms of food needs in terms of nesting habits physiological response voltinism Etc and this can lead to differences in responses to environmental gradients for example and potentially various population Trends with this people we end that have having a national scale coverage because France is very diverse in terms of Bio biogeographical regions and also in terms of habitats here I only show three broad categories that I will reuse again after Urban Agricultural and natural but anyway it’s very diverse in terms of habitat so it’s important to have data from from everywhere if possible so to sum up with this project we aim at having uh obtaining data on a broad teic scope keeping the interaction with the plant not just the insect but also the plant that the insects visit a national scale if possible long-term survey in order to in the end get population Trends and um having a non Leal protocol as well for the observations and this is done since 2010 by citizen scientist so how does it work exactly in other words what are they doing here so it’s based on citizen science so very briefly the involvement of Volunteers in scientific research so it’s a a collaboration between a scientific team and a community of volunteers working around on the same questions so it’s quite simple but it’s yet standardized anywhere in France uh participant will pick fling plant species of Interest the one he wants or she wants and then the aim is to photograph all insects that visit the flowers of that species so in the end you obtain a picture of the plants and the set of visitors that visited the plants during the the session at first in 20110 we had two sampling effort options we had a a flash option which is 20 20 minute exactly of photographic session but we also allowed for a longer session which so we call it a long protocol and so it allowed the participant to make observation during more than 20 minutes and also to come back to the flower the species uh multiple times during three days but since 2010 2018 sorry we uh stopped the long protocol just to increase the number of exactly standardized protocol and making sure we have more and more 20 minute exactly sampling effort um after this session of photographing the participant hands up with a lot of pictures of insects and the aim is to sort them into moros species there is a huge diversity of flower visitors and identify identifying them is very difficult as Den just said for example for this one you have to capture it kill it pin it uh view it under a microscope and look at a very specific detail and we chose a different approach so using morphos species defined as a group of species that differs from the other morphos species in any external features that can be seen on pictures of a live insect to have this approach we developed an interactive online tool so it’s just like a key but it’s not diatomic here on the left you have the different uh descriptors of the morphos species so different criteria and uh people can answer them uh depending on what they can see on their pictures and um they don’t have to answer all of them but if you answer them all you are more and more um you get more chance to to to end up with a particular morphos species and to finalize your identification this is based on the expert three um tool that makes a different kind of interactive tools so about the moros species just some some information on their taxonomic resolution because as you can imagine it’s not all species because it’s based on pictures so here you have the number of defined morphos species in the online tool the identification key and uh here you have the number of defined moros species by tax resolution so it goes from one for a group of species from different families within World Order and until seven for single species so species uh to the um strictly speaking so for hopa we have still a fair amount of species but many species many morphos species sorry are not defined at the species level um the same is true for the other uh main orders of of flower visitors and uh the order that is a little bit different is for butterflies because many species can be potentially identified on pictures if you have enough pictures and potentially from both sides of the Wings in total we have 630 moros species defined so that’s uh grouping all all flower visitors in France so they are all put into the 650 30 different MOS species and almost half of them are species strictu and for 70% of them the genius is none for plants we are using now plant net so I’m not going to say more about it because it’s the next talk and how it works uh after that so people um put online their pictures and there is a need to validate the identification for example for this picture the morphos she is the black and yellow slies and that was proposed by the auor of the of the picture so the participant and in 2010 participants could comment on that picture and even emit D so they would flag that picture like maybe this is not sure and entomologist from the op would validate the the identification validate or change the identification so that was a lot of job of work for these people and that was changed for different reason and since 2018 19 the what we call peer validated peer validation uh possibilities uh it took a lot of time for experts to validate all the data also the participant gained a lot of entomological skills with participation and there was also a will to allow for new way of participating to the project so not just collecting data but also validating the data for those who are capable of it and so now people or participants they can validate the identification and if there are more than three valid or if there are three validation then we consider that identification to be uh validated and uh approximately we have more than 95% of pictures that have received at least three validation uh by the other participants so it works pretty well and they’re really happy to be able to contribute that so to sum up we have a data on interactions between plants and visitors we have a location we have a sampling effort date and time obviously weather condition and all of that is what we call the collection in in those people and this is how where are distributed in France so we have more than 7 70,000 collections in France since 2010 so some main results um so we have two kind of of results we started by analyzing the responses of FL visitor to land uses and especially urbanization and more recently we we had some work on the consequences of participating for volunteers so more uh social let’s say research but I’m not going to talk about this which show today research today very basically what we did was to combine data on so pollinator data with land use data and then we estimated the response of the changes in occurrences for different morphos species according to different L use types we computed L use Affinity indexes so indicators of affinity for the urban agricultural or natural L use types for our flower Vos and the idea was to answer that question in which L use are found a wild flower visitors at the scale of France and R CS refugees for example so here on the y axis you have the urban L use affinity and each dot here is a morphos species of the other hopa for those that are below zero it’s a an urban avoider moros species so it’s less seen in urban areas than expected by chance if you have if a moros species is above the line it’s what we call colban exploiters and if it’s overlapping with zero either there are tolerance so we don’t have enough data to know for sure so if you look at hopa in aage it looks like this order is rather tolerant to urbanization but if we consider now the other orders we have much more urban avoider spe moros species in these orders in 2010 when I did the analyzis we had about 42% of the morphos species analyze that were ban avoiders and recently I uh I redone this analysis with data from 2010 to 2022 so I could analyze 235 morphos species and out of these we had 72% of morphos species that were significantly Urban voiders so if you sum up this graph it’s here again so it’s the Affinity here for urban L use for the four orders be flies hopa and butterflies and now I’m going to show you the affinity for the agrial land use and urban l jues so of course it’s a a bit different the mean average affinity for agricultural L use is um rather positive so it’s to me it’s not surprising I mean it’s more interesting for flors because there are more resources than in urban areas also again for hopter aage we don’t have a significant signal and for natural L use it’s particularly positive for bitles and for butterflies now on these graphs the black dots are actually only the what we could call the common morphos species so common in the sense that we recorded them often in our observations of those people so common means that they were observed in more than 2% of the collections made in at least one of the three L Type and otherwise there are infrequent moros species that means that they were observed in less than 2% of the collections made in each of the three land ju types so this was made to correct for the BIOS we have in the sense that we have more collection made in urban areas for example so now we can see how they differ in their affinity for the different L types and for their Affinity with Urban L the infrequent moros species are more urban avoiders they avoid even more urban habitat than the common moros species there is no difference regarding the affinity for agricultural land use and there is a difference for natural land use so the infrequent morphos species tend to have a more positive affinity for natural L ju and given the definition of the moros species the infrequent moros species include more species in the taxonomic sense than the common morphos species one last results very briefly here you have a an analyze again on the three L use types but the three L use type taken together and the interactions the triple interactions between them so the result will be shown on this uh tary plot and you have here Urban proportion in the landscape agrial proportion in the landscape and seminatural proportion in the landscape so landscape here means that we are in a 20% Urban landscape 10% agal area and 70% seminatural areas and the color here in the tary plot is the predicted richness well number of morphos species within 20 minutes um based on the collections made by the participants so what we can see is that the maximum predicted richness is uh in Landscapes with about 70% agricultural areas and 30 40% natural areas and on the contrary urban areas um R po compared to the other Landscapes and actually wherever you are on the stary plot so in where whichever landscape you are if you increase the urbanization it decreases the predicted richness of flower visitors that you will get in your collection so this paper goes into more details of well the same plots but for different groups of moros species but I’m not saying more right now so one might wonder so why why do they vary in their response to urbanization so it can be there can be many different reasons for the variable response to urbanization and something we looked at is the specialization toward floral resources uh so on this graph here you have the mean Community specialization so the mean specialization in the collection made by the participants if you are at the top it’s a collection made of res special species and at the bottom ra generate species and we are going to see what happens with increasing proportion of fan L use and there is a decreasing Trend so that means that there is more and more generous species in frequency in the collections made by participant in urban areas along the urban gradient and this um correspond to a functional biotic homogen homogenization with urbanization at the scale of FR so some scientific perspectives that we have uh there is uh currently an ongoing study on the effect of pesticides on as much as possible flower visitors that we can study with those people so we will see uh so this is BAS based sorry on a new database of annual pesticide sales in France at the at at a quite fine scale geographical scale because it’s a city or a few City and it’s annal and it’s the sales for each substance pesticide substance so it’s quite precise and if we link this with the toxicity of the substances half life the metabolites Etc then we came up with a index which we call for now a total Hazard ratio but we can also have information on a sort of a cocktail index because we have all these hundreds and hundreds of substances uh I don’t have any results yet unfortunately but what I can say is that some models are running for about 2040 species that we are able to identify on pictures uh some interesting perspectives will be as I said at the beginning to try to get some population trends um so we could try to begin to do that because we have 14 years of data over 600,000 okay plan visitor interaction records more than 4,000 participants so that looks pretty great but there is still some challenges actually only a small core of participants are long-term participants and that’s an issue with most citizen science project unfortunately but for example we have about 4% of our participants that stay for more than 5 years um and there is this High turnover of participants so 85% of them stay for only one year so that’s a huge Challenge and that’s uh that’s the case even though we have newsletter annual meetings webinars and a lot of things going on so we still have to work on that uh so about population trains I I have nothing more to say expect except that it’s a perspective it it needs to be assessed whether or not we can do it with some certainties uh something that would be interesting is to actually get to the interaction networks level we uh potentially can do it because we have the insects interacting with the plants and in this paper by EST Reno from the University of by she investigated um plant pator networks available in the literature and defined at the species level and she tried to see how to uh what was the influence of downgraded D grading them that mean lowering the taxonomic resolution of the insects and also of the plants and also together how that would impact the Matrix of the interaction Networks so in terms of metrics she looked at nestin connectance modularity Etc so different classic metrics looked that uh in interaction interaction networks and the to very sum up to to to sum up her findings the relative values of the different indicators remain the same never um regardless the taxonomic resolution so that means that if we stick to one taxonomic resolution we can rank the different networks and this is interesting because applied to those people we could build interaction networks plan visitor networks at the the people taxonomic resolution and assess how the metrics that we compute maybe change or not according to different environmental gradients such as urbanization pesticides different agrial practices uh also with time perhaps so again some perspectives and finally that’s perhaps a question or a dream I don’t know for me but we could wonder if it can be expended to other European countries in France we we well we have four biogeographical regions and these are present in most of Europe as well uh but I suppose that’s not enough because uh we should still make probably a lot of updates in the interactive online tools to include for new species adid as there will be new species in Spain or in other countries so there will be a lot of work to to be expanding I suppose and the same for plants although we have plant net so see next talk I like I would like to thank you for attention again the participants and everyone who contributed to to those people [Applause] [Music]